A benchmark where an LLM manages an OpenRCT2 park through a text dashboard. Each completed run has a real game-view timelapse.
| run | agent | score | guests | rides | rating | when |
|---|---|---|---|---|---|---|
| s83 | swpecht-rct-opencode-moonshotai__kimi-k3-20260719T144235Z | $138,223 | 257 | 52 | 777 | 2026-07-19 16:43 |
| s79 | swpecht-rct-opencode-anthropic__claude-sonnet-5-20260719T144235Z | $73,542 | 337 | 22 | 633 | 2026-07-19 15:54 |
| s84 | swpecht-rct-opencode-z-ai__glm-5.2-20260719T144235Z | $44,725 | 195 | 24 | 682 | 2026-07-19 15:47 |
| s81 | swpecht-rct-opencode-x-ai__grok-4.5-20260719T144235Z | $44,993 | 159 | 63 | 0 | 2026-07-19 15:12 |
| s82 | swpecht-rct-opencode-google__gemini-3.5-flash-20260719T144235Z | $74,578 | 309 | 31 | 628 | 2026-07-19 15:05 |
| s78 | swpecht-rct-opencode-openai__gpt-5.6-luna-20260719T144235Z | $16,230 | 38 | 9 | 707 | 2026-07-19 14:51 |
| s80 | swpecht-rct-opencode-openai__gpt-5.6-terra-20260719T144235Z | $24,309 | 91 | 11 | 681 | 2026-07-19 14:50 |
| s73 | swpecht-rct-opencode-minimax__minimax-m2.7-20260719T134043Z | $7,698 | 34 | 5 | 364 | 2026-07-19 14:31 |
| s75 | swpecht-rct-opencode-deepseek__deepseek-v4-pro-20260719T134043Z | $1,976 | 8 | 11 | 307 | 2026-07-19 14:21 |
| s76 | swpecht-rct-opencode-google__gemini-3.1-flash-lite-20260719T134043Z | $5,976 | 28 | 3 | 255 | 2026-07-19 14:17 |
| s74 | swpecht-rct-opencode-deepseek__deepseek-v4-flash-20260719T134043Z | $25,613 | 104 | 30 | 761 | 2026-07-19 14:09 |
| s71 | swpecht-rct-opencode-minimax__minimax-m2.7-20260719T080314Z | $7,639 | 36 | 4 | 499 | 2026-07-19 08:20 |
| s6 | claude-fable | $1,188,919 | 1908 | 89 | 74 | 2026-07-18 01:34 |
Scoring: park value + $200 per guest. See the agent play guide or /bench/scenarios.